From Impersonation to Delegation.

Agentic commerce is expected to make the fastest leap yet from emergence to mainstream adoption.

Gartner forecasts that by 2028, AI agents will drive 20% of digital storefront interactions and make 15% of day-to-day work decisions autonomously. To capitalise on this shift, organisations need delegation models and fine-grained authorisation & entitlement systems that are fit for purpose.

The problem? Today’s digital services were built for humans—not autonomous agents. The first wave of agentic innovation has relied on impersonation, with AI agents mimicking users to navigate interfaces, without formally declared identities or clearly defined authority. It’s sophisticated screen scraping at best.

This lack of transparency increases the risk of unintended actions, misused data, and regulatory breaches. As MIT highlights:

“The rapid deployment of autonomous AI agents creates urgent challenges around authorisation, accountability, and access control in digital spaces.”

To meet this challenge we need to move from implicit trust to explicit delegation, grounded in enforceable policies and accountability at the level of the agent. This is the foundation of knowing your agent— KYA.

This transition must be underpinned by Digital Public Infrastructure (DPI)—the foundational layers of digital identity, payments, and data exchange that enable safe participation in the digital economy. As the World Economic Forum puts it:

“DPI provides the shared digital rails upon which innovation can occur, enabling both public and private actors to build interoperable solutions that benefit society at large.”

In this model, governments play an important role in establishing the root of trust and policy guardrails, while the private sector builds scalable, cross-sector solutions. DPI is not a single platform or technology; it is “an ecosystem of enabling digital systems” where interoperability and trust are paramount.